Repository navigation
Commit 193b808
pythongh-146306: Specialize float/float true division in tier 2 optimizer
Float/float true division is not specialized at tier 1, going through
generic BINARY_OP dispatch. This adds tier 2 specialization that the
optimizer emits when at least one operand is known to be a float:
- _BINARY_OP_TRUEDIV_FLOAT (non-inplace)
- _BINARY_OP_TRUEDIV_FLOAT_INPLACE (unique LHS)
- _BINARY_OP_TRUEDIV_FLOAT_INPLACE_RIGHT (unique RHS)
The optimizer inserts _GUARD_TOS_FLOAT / _GUARD_NOS_FLOAT for
operands not yet proven to be float, enabling specialization in
expressions like `(a + b) / c` where only the LHS type is known.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>1 parent 6009309 commit 193b808
7 files changed
Lines changed: 1641 additions & 1218 deletions
File tree
- Include/internal
- Python
0 commit comments